Comment Control of Nanobots (Score 1) 178
As the article points out, the key to using nanotechnology effectively is self-replication. It is too difficult to manufacture the millions of nanobots that would be required to complete any one task. Essentially, nanotechnology should work by humans designing and making a single (or a few) nanobots which are designed to replicate under certain conditions. From there, the nanobot works as a Von Neumann machine, replicating itself from raw materials provided and continually working on the task it's been made to complete, whether it be microsurgery or buckyball construction or whatever. However, there is the problem of limiting replication... one must consider the possible military applications of this technology.
There are ways of limiting replication, however. In the cells of animals (and maybe plants - I'm not too sure about that, though), there are telomeres, short dna fragments that do not replicate when the dna cell does, halving in size with each cell division, eventually inhibiting division when it reaches a certain size. This idea could be applied to nanobots, easily limiting replication to a certain limit (say one hundred or so - 2^100 nanobots should be able to accomplish most any task). However, this must be implemented by the creators of the nanobots - which raises the question of whether or not implementation of this is possible.
As most any programmer knows, programming embedded systems is often quite difficult due to the limited ram in most of those devices. Creating a nanobot that is both able to replicate and able to complete the task it is designed for would require quite complex programming, likely beyond that which can be accomplished through the use of straightforward electrical circuits. Some sort of processor with instructions for that processor would be required. Creating a processor that small with a flash rom for machine code and ram for variables could prove to be impossible. An entirely new method for control of these nanobots would have to be developed if we are to effectively use them.
However, this method is many, many years off. The only feasible use for nanotechnology at this point is for converting some simple matter into other simple matter. Perhaps mining certain rare metals from ore could be done in the next ten or twenty years. Medical nanobots and weaponry are far, far off in the future. (though you can never underestimate human ingenuity - it has only been forty or fifty years since computers were first made and now they run most everything)
There are ways of limiting replication, however. In the cells of animals (and maybe plants - I'm not too sure about that, though), there are telomeres, short dna fragments that do not replicate when the dna cell does, halving in size with each cell division, eventually inhibiting division when it reaches a certain size. This idea could be applied to nanobots, easily limiting replication to a certain limit (say one hundred or so - 2^100 nanobots should be able to accomplish most any task). However, this must be implemented by the creators of the nanobots - which raises the question of whether or not implementation of this is possible.
As most any programmer knows, programming embedded systems is often quite difficult due to the limited ram in most of those devices. Creating a nanobot that is both able to replicate and able to complete the task it is designed for would require quite complex programming, likely beyond that which can be accomplished through the use of straightforward electrical circuits. Some sort of processor with instructions for that processor would be required. Creating a processor that small with a flash rom for machine code and ram for variables could prove to be impossible. An entirely new method for control of these nanobots would have to be developed if we are to effectively use them.
However, this method is many, many years off. The only feasible use for nanotechnology at this point is for converting some simple matter into other simple matter. Perhaps mining certain rare metals from ore could be done in the next ten or twenty years. Medical nanobots and weaponry are far, far off in the future. (though you can never underestimate human ingenuity - it has only been forty or fifty years since computers were first made and now they run most everything)
